@@ -128,104 +128,92 @@ def inpxml_todict(parent: etree._Element,
128128
129129 :return: a python dictionary
130130 """
131+ #These keys have to never appear as an attribute/tag name
132+ #The underscores should guarantee that
133+ _TEXT_PLACEHOLDER = '__text__'
134+ _OMIT_PLACEHOLDER = '__omit__'
131135
132136 #Check if this is the first call to this routine
133137 if base_xpath is None :
134138 base_xpath = f'/{ parent .tag }
135139
136- return_dict : dict [str , Any ] = {}
137- if list (parent .items ()):
138- return_dict = {str (key ): val for key , val in parent .items ()}
139- # Now we have to convert lazy fortran style into pretty things for the Database
140- for key in return_dict :
141- if key in schema_dict ['attrib_types' ]:
142- return_dict [key ], suc = convert_from_xml (return_dict [key ],
140+ content : dict [str , Any ] = {}
141+ # Now we have to convert lazy fortran style into pretty things for the Database
142+ for key , value in parent .items ():
143+ attrib_name , value = str (key ), str (value )
144+ if attrib_name in schema_dict ['attrib_types' ]:
145+ content [attrib_name ], suc = convert_from_xml (value ,
143146 schema_dict ,
144- key ,
147+ attrib_name ,
145148 text = False ,
146149 constants = constants ,
147150 logger = logger )
148- if not suc and logger is not None :
149- logger .warning ("Failed to convert attribute '%s' Got: '%s'" , key , return_dict [key ])
150-
151- if parent .text :
152- # has text, but we don't want all the '\n' s and empty strings in the database
153- if parent .text .strip () != '' : # might not be the best solutions
154- if parent .tag not in schema_dict ['text_tags' ]:
155- if logger is not None :
156- logger .error ('Something is wrong in the schema_dict: %s is not in text_tags, but it has text' ,
157- parent .tag )
158- raise ValueError (
159- f'Something is wrong in the schema_dict: { parent .tag } )
160-
161- converted_text , suc = convert_from_xml (str (parent .text ),
162- schema_dict ,
163- parent .tag ,
164- text = True ,
165- constants = constants ,
166- logger = logger )
167-
168151 if not suc and logger is not None :
169- logger .warning ("Failed to text of '%s' Got: '%s'" , parent .tag , parent .text )
152+ logger .warning ("Failed to convert attribute '%s' Got: '%s'" , attrib_name , value )
153+
154+ # has text, but we don't want all the '\n' s and empty strings in the database
155+ if parent .text and parent .text .strip () != '' :
156+
157+ if parent .tag not in schema_dict ['text_tags' ]:
158+ if logger is not None :
159+ logger .error ('Something is wrong in the schema_dict: %s is not in text_tags, but it has text' ,
160+ parent .tag )
161+ raise ValueError (
162+ f'Something is wrong in the schema_dict: { parent .tag } )
170163
171- if not return_dict :
172- return_dict = converted_text #type:ignore
173- else :
174- return_dict ['text_value' ] = converted_text
175- if 'label' in return_dict :
176- return_dict ['text_label' ] = return_dict ['label' ]
177- return_dict .pop ('label' )
164+ converted_text , suc = convert_from_xml (str (parent .text ),
165+ schema_dict ,
166+ parent .tag ,
167+ text = True ,
168+ constants = constants ,
169+ logger = logger )
170+
171+ if not suc and logger is not None :
172+ logger .warning ("Failed to text of '%s' Got: '%s'" , parent .tag , parent .text )
173+
174+ content [_TEXT_PLACEHOLDER ] = converted_text
178175
179176 tag_info = schema_dict ['tag_info' ].get (base_xpath , EMPTY_TAG_INFO )
180177 for element in parent :
181178
182- new_base_xpath = f'{ base_xpath } { element .tag }
183- omitt_contained_tags = element .tag in schema_dict ['omitt_contained_tags' ]
184- new_return_dict = inpxml_todict (element ,
185- schema_dict ,
186- constants ,
187- base_xpath = new_base_xpath ,
188- omitted_tags = omitt_contained_tags ,
189- logger = logger )
190-
191- if element .tag in tag_info ['several' ]:
192- # make a list, otherwise the tag will be overwritten in the dict
193- if element .tag not in return_dict : # is this the first occurrence?
194- if omitted_tags :
195- if len (return_dict ) == 0 :
196- return_dict = [] #type:ignore
197- else :
198- return_dict [element .tag ] = []
199- if omitted_tags :
200- return_dict .append (new_return_dict ) #type:ignore
201- elif 'text_value' in new_return_dict :
202- for key , value in new_return_dict .items ():
203- if key == 'text_value' :
204- return_dict [element .tag ].append (value )
205- elif key == 'text_label' :
206- if 'labels' not in return_dict :
207- return_dict ['labels' ] = {}
208- return_dict ['labels' ][value ] = new_return_dict ['text_value' ]
209- else :
210- if key not in return_dict :
211- return_dict [key ] = []
212- elif not isinstance (return_dict [key ], list ): #Key seems to be defined already
213- if logger is not None :
214- logger .error ('%s cannot be extracted to the next level' , key )
215- raise ValueError (f'{ key } )
216- return_dict [key ].append (value )
217- for key in new_return_dict .keys ():
218- if key in ['text_value' , 'text_label' ]:
219- continue
220- if len (return_dict [key ]) != len (return_dict [element .tag ]):
179+ child_content = inpxml_todict (element ,
180+ schema_dict ,
181+ constants ,
182+ base_xpath = f'{ base_xpath } { element .tag } ,
183+ omitted_tags = element .tag in schema_dict ['omitt_contained_tags' ],
184+ logger = logger )
185+
186+ if _OMIT_PLACEHOLDER in child_content :
187+ #We knoe that there is only one key here
188+ child_content = child_content .pop (_OMIT_PLACEHOLDER )
189+
190+ tag_name = element .tag
191+ if omitted_tags :
192+ tag_name = _OMIT_PLACEHOLDER
193+
194+ if element .tag in tag_info ['several' ]\
195+ and _TEXT_PLACEHOLDER in child_content :
196+ #The text is stored under the name of the tag
197+ text_value = child_content .pop (_TEXT_PLACEHOLDER )
198+ content .setdefault (tag_name , []).append (text_value )
199+ child_tag_info = schema_dict ['tag_info' ].get (f'{ base_xpath } { element .tag } , EMPTY_TAG_INFO )
200+ for key , value in child_content .items ():
201+ if key not in child_tag_info ['optional_attribs' ]:
202+ #All required attributes are stored as lists
203+ if key in content and \
204+ not isinstance (content [key ], list ): #Key seems to be defined already
221205 if logger is not None :
222- logger .error (
223- 'Extracted optional argument %s at the moment only label is supported correctly' , key )
224- raise ValueError (
225- f'Extracted optional argument { key } )
226- else :
227- return_dict [element .tag ].append (new_return_dict )
206+ logger .error ('%s cannot be extracted to the next level' , key )
207+ raise ValueError (f'{ key } )
208+ content .setdefault (key , []).append (value )
209+ else :
210+ #All optional attributes are stored as dicts pointing to the text
211+ content .setdefault (key , {})[value ] = text_value
212+ elif element .tag in tag_info ['several' ]:
213+ content .setdefault (tag_name , []).append (child_content )
214+ elif _TEXT_PLACEHOLDER in child_content :
215+ content [tag_name ] = child_content .pop (_TEXT_PLACEHOLDER )
228216 else :
229- return_dict [ element . tag ] = new_return_dict
217+ content [ tag_name ] = child_content
230218
231- return return_dict
219+ return content
0 commit comments